PROCEEDINGS OF THE SOCIETY FOR EXPERIMENTAL BIOLOGY AND MEDICINE
150. 134-135 (1975)
Experimental Pneumococcal Meningitis: 111. Chemotactic Activity in Cerebrospinal
Fluid' (38989) CHARLES M. NOLAN,2 ROBERT A. CLARK, AND HARRY N. BEATY (Introduced by R . G. Petersdorf) Department of Medicine, University of Washington and Providence Medical Center, Seattle, Washington 98122
Evidence is accumulating that mortality in pneumococcal meningitis is related t o inflammation in and around the meninges (1-3). Quantitative determinations of meningeal inflammation in a rabbit model of pneumococcal meningitis demonstrated an increase in leukocyte infiltration early in the infection with the peak at about the time the average animal dies (3): This study was undertaken t o determine the stimulus for this influx of acute inflammatory cells. Neutrophil chemotactic activity was detected in cerebrospinal fluid (CSF) in the animal model of meningitis and the responsible factors were partially characterized. Methods. The method of inducing pneumococcal meningitis in rabbits (intracisternal mucin plus iv pneumococci, Type 111) and the clinical course of the infection have been described in detail (4). Groups of animals infected simultaneously were subjected t o cisternal puncture either 24, 48, or 72 hr after infection. Supernatants of CSF from several infected rabbits were pooled, passed through a bacterial filter and tested undiluted for chemotactic activity for normal circulating rabbit granulocytes. Pooled CSF was concentrated five to tenfold by ultrafiltration (Amicon Model 12 with UM-2 filter, Amicon Corp., Lexington, MA) prior to fractionation by gel filtration on Calibrated columns of Sephadex G-25 and G-75 (Pharmacia Fine Chemicals, Piscataway, NJ). Fractions of 2 ml eluted with 0.01 M This work was supported by Grants CC00665 and A103456 from the Center for Disease Control and the National Institute of Allergy and Infectious Disease. 2Address reprint requests to Charles M. Nolan, M.D., Department of Medicine, University of Arkansas School of Medicine, Little Rock, Arkansas 72201.
phosphate buffered normal saline at 4" were assayed for chemotactic activity and protein content (absorbance at 280 mn). Individual 72-hr CSF specimens were analyzed for total hemolytic complement (sensitized sheep erythrocyte hemolysis) and C4 (radial immunodiffusion against antibody to human C4, Meloy Laboratories, Springfield, VA). CSF and concentrated chemotactically active fractions from a Sephadex G-75 column were absorbed with antihuman C3 and C5 sera (Meloy Laboratories) or antitype I11 pneumococcal serum (Difco Labs., Detroit, MI) and then tested for chemotactic activity. Purified Type I11 pneumococcal polysaccharide (courtesy of Dr. George Kenney) at 10 and 50 pg/ml and supernatants of cultures of Type I11 pneumococci in medium 199 (BBL, Cockeysville, MD) were also tested for chemotactic activity. Chemotaxis was measured by a previously described method ( 5 ) employing two-compartment chambers (Ahlco Machine Company, New Brunswick, CN, and Mark-it Corp., Chicago, IL) and rabbit blood granulocytes prepared by dextran sedimentation and labeled with 5 1 C h r ~ m i ~(6). m Results were expressed as mean corrected CPM ( 5 ) of triplicate chemotaxis chambers =tSEM. Results. Control CSF from noninfected rabbits had no granulocyte chemotactic activity (mean of 16 experiments 156 CPM =I= 17 versus a mean for Gey's medium (6) controls of 234 CPM j = 51 in 8 experiments). In contrast, a progressive increase in chemotactic activity of CSF was observed following the development of meningitis (Fig. 1). The activity at 48 and 72 hr was consistently and significantly (P < .Ol) greater than control values. Rabbits given intracisternal mucin but no iv pneumococci developed a transient chemical meningitis
134 Copyright @ 1975 by the Society for Experimental Biology and Medicine All rights reserved.
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
135
CSF CHEMOTACTIC ACTIVITY IN MENINGITIS
Fraction
I
Control
24
,
48
I
72
Time ( h o u r s )
number
FIG.2. Sephadex G-75gel filtration of CSF from rabbits with pneumococcal meningitis. Absorbance at 280 nm (-) and chemotactic activity (stippled bars) of each fraction are plotted. Arrows indicate void volume (Vo) and elution position of marker proteins.
FIG. 1. Chemotactic activity in CSF of rabbits with pneumococcal meningitis. Each point represents the mean f SEM of the number of experiments indicated. *Denotes values significantly different from controls (P < .01, t test).
nor C4 were detectable in infected CSF. Absorption of infected CSF or concentrated fractions comprising the 11,000 mol wt peak of chemotactic activity from a Sephadex G-75 column with antihuman C3 and C5 (4) and their CSF contained moderate sera did not reduce chemotactic activity. chemotactic activity after 24 hr but none at Approximately 50 % of the chemotactic activity was removed from an aliquot of 72 hr. infected CSF by absorption with antiserum Chemotactic activity in infected CSF was not dialyzable and was stable at -80" fclr to type I11 pneumococci. Significant chemo2 weeks, 4" for 1 week and 56" for 30 min. tactic activity was detected in filtrates of Activity was abolished by heating at 100" overnight cultures of type 111 pneumococci for 15 min. Supernatants of CSF pre- and in solutions of purified Type 111 pneucipitated with 5 % trichloracetic acid and mococcal capsular polysaccharide. Discussion. This study, showing that granboth supernatants and redissolved precipiulocyte chemotactic activity appears in CSF tates of CSF saturated with ammonium of rabbits with pneumococcal meningitis, sulfate were devoid of chemotactic activity. adds another to the list of chemotacticallyChromatography of infected CSF on active biologic fluids from areas of inflamSephadex G-25 resulted in elution of most of the protein in the exclusion volume (mol mation (8-11). Note that the time course of wt greater than 10,000). Chemotactic ac- appearance of chemotactic activity in CSF tivity was also concentrated in the exclusion during infection closely parallels that of volume, but a small amount of activity was accumulation of the inflammatory mass of present in fractions eluting after the marker granulocytes within the meninges (3). A cause-and-effect relationship between protein insulin, On a Sephadex G-75 column these two observations has not been estab(Fig. 2) most protein again eluted with the lished. Nevertheless, the detection of a exclusion volume (mol wt > 50,000). Howpotential mediator of central nervous system ever, the major peak of chemotactic activity occurred in fractions eluting between ribo- inflammation assumes importance in view nuclease and insulin, corresponding t o a of increasing evidence that meningeal inmol wt of about 11,000; a smaller peak flammation is a determinant of mortality in eluted after insulin, indicating a mol wt of pneumococcal meningitis. If this supposition proves correct in subsequent experiments, approximately 3000. Neither hemolytic complement activity suppression of the inflammatory response
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
136
CSF CHEMOTACTIC ACITVITY IN MENINGITIS
might be beneficial in the treatment of creasing levels for 72 hr after infection. pneumococcal meningitis. Modification of Activity was stable at 56” and was inactivated the chemotactic stimulus is one possible by agents which denature proteins. Gel therapeutic approach t o this problem; char- filtration demonstrated two chemotactically acterization of CSF chemotactic factors active fractions in infected CSF with mol wts and their origin would facilitate this ap- of approximately 3000 and 11,000. Bacterial proac h. products appear to account for a portion of Our preliminary work suggests that CSF the observed CSF chemotactic activity, but chemotactic activity is attributable to rela- the role of host factors remains to be clarified. tively low molecular weight substances which are heat stable and either protein in nature 1. Beaty, H. N., and Oppenheimer, S., N. Eng. J. or dependent upon protein for their activity. Med. 279, 1197 (1968). 2. Sears, M. R., O’Donoghue, J. M., Fisher, H. K., Chemotactic activity of infected CSF was and Beaty, H. N., J. Clin. Invest. 54, 18 (1974). party absorbed by specific pneumococcal 3. McAllister, C. K., O’Donoghue, J. M., and antiserum, suggesting that a bacterial factor Beaty, H. N., J. Infect. Dis. Vol. 132 (1975). accounts for a portion of it. Pneumococci produce low molecular weight ( < 3600) 4. O’Donoghue, J. M., Schweid, A. I., and Beaty, H. N., Proc.Soc.Exp. Biol. Med. 146, 571 (1974). chemotactic activity during growth (12). 5. Gallin, J. I., Clark, R. A., and Kimball, H. R., J. Release of complement derived chemoImmunol. 110, 233 (1973). tactic factors in infected CSF might also be 6. Clark, R. A., and Kimball, H. R., J. Clin. Invest. expected since pneumococci activate com50, 2645 (1971). plement via the alternate pathway (13). 7. Green, A. A. and Hughes, W. L., in “Methods in Although complement activation products Enzymology (S.R. Colowick and N. 0. Kaplan, eds.), Vol. I, p. 67. Academic Press, New York cannot be implicated on the basis of current (1955). evidence, CSF chemotactic activity detected 8. Snyderman, R., Phillips, J. K., and Mergenhagen, in this model has characteristics in common S. E., J. Exp. Med. 134, 1131 (1971). with the potent chemotactic agent C5a. 9. Ward, P. A., and Zwaifler, N. J., J. Clin. Invest. Both are stable at 56” and the mol wt of 50, 606 (1971). rabbit C5a (12,500) (14) is similar to that of 10. Hill, J. H., and Ward, P. A., J. Exp. Med. 133, the major CSF chemotactic peak separated 885 (1971). by gel filtration. Clarification of the roles of 11. Ward, P. A,, and Hill, J. H., J. Immunol. 108, microbial products and host factors in the 1137 (1972). inflammatory response in meningitis is criti- 12. Ward, P. A., Lepow, J. H., and Newman, L. J., Amer. J. Pathol. 52, 725 (1968). cal and deserves further study. Summary. Chemotactic activity was as- 13. Winkelstein, J. A., Shin, H. S., and Wood, W. B., Jr., J. Immunol. 108, 1681 (1972). sayed in CSF of rabbits with pneumococcal 14. Snyderman, R., Phillips, J., and Mergenhagen, meningitis to further characterize the inS. E., Infect. Immunol. 1, 521 (1970). flammatory response in this infection. CSF chemotactic activity was detected in in- Received May 6, 1975. P.S.E.B.M. 1975, Vol. 150.
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
150. 134-135 (1975)
Experimental Pneumococcal Meningitis: 111. Chemotactic Activity in Cerebrospinal
Fluid' (38989) CHARLES M. NOLAN,2 ROBERT A. CLARK, AND HARRY N. BEATY (Introduced by R . G. Petersdorf) Department of Medicine, University of Washington and Providence Medical Center, Seattle, Washington 98122
Evidence is accumulating that mortality in pneumococcal meningitis is related t o inflammation in and around the meninges (1-3). Quantitative determinations of meningeal inflammation in a rabbit model of pneumococcal meningitis demonstrated an increase in leukocyte infiltration early in the infection with the peak at about the time the average animal dies (3): This study was undertaken t o determine the stimulus for this influx of acute inflammatory cells. Neutrophil chemotactic activity was detected in cerebrospinal fluid (CSF) in the animal model of meningitis and the responsible factors were partially characterized. Methods. The method of inducing pneumococcal meningitis in rabbits (intracisternal mucin plus iv pneumococci, Type 111) and the clinical course of the infection have been described in detail (4). Groups of animals infected simultaneously were subjected t o cisternal puncture either 24, 48, or 72 hr after infection. Supernatants of CSF from several infected rabbits were pooled, passed through a bacterial filter and tested undiluted for chemotactic activity for normal circulating rabbit granulocytes. Pooled CSF was concentrated five to tenfold by ultrafiltration (Amicon Model 12 with UM-2 filter, Amicon Corp., Lexington, MA) prior to fractionation by gel filtration on Calibrated columns of Sephadex G-25 and G-75 (Pharmacia Fine Chemicals, Piscataway, NJ). Fractions of 2 ml eluted with 0.01 M This work was supported by Grants CC00665 and A103456 from the Center for Disease Control and the National Institute of Allergy and Infectious Disease. 2Address reprint requests to Charles M. Nolan, M.D., Department of Medicine, University of Arkansas School of Medicine, Little Rock, Arkansas 72201.
phosphate buffered normal saline at 4" were assayed for chemotactic activity and protein content (absorbance at 280 mn). Individual 72-hr CSF specimens were analyzed for total hemolytic complement (sensitized sheep erythrocyte hemolysis) and C4 (radial immunodiffusion against antibody to human C4, Meloy Laboratories, Springfield, VA). CSF and concentrated chemotactically active fractions from a Sephadex G-75 column were absorbed with antihuman C3 and C5 sera (Meloy Laboratories) or antitype I11 pneumococcal serum (Difco Labs., Detroit, MI) and then tested for chemotactic activity. Purified Type I11 pneumococcal polysaccharide (courtesy of Dr. George Kenney) at 10 and 50 pg/ml and supernatants of cultures of Type I11 pneumococci in medium 199 (BBL, Cockeysville, MD) were also tested for chemotactic activity. Chemotaxis was measured by a previously described method ( 5 ) employing two-compartment chambers (Ahlco Machine Company, New Brunswick, CN, and Mark-it Corp., Chicago, IL) and rabbit blood granulocytes prepared by dextran sedimentation and labeled with 5 1 C h r ~ m i ~(6). m Results were expressed as mean corrected CPM ( 5 ) of triplicate chemotaxis chambers =tSEM. Results. Control CSF from noninfected rabbits had no granulocyte chemotactic activity (mean of 16 experiments 156 CPM =I= 17 versus a mean for Gey's medium (6) controls of 234 CPM j = 51 in 8 experiments). In contrast, a progressive increase in chemotactic activity of CSF was observed following the development of meningitis (Fig. 1). The activity at 48 and 72 hr was consistently and significantly (P < .Ol) greater than control values. Rabbits given intracisternal mucin but no iv pneumococci developed a transient chemical meningitis
134 Copyright @ 1975 by the Society for Experimental Biology and Medicine All rights reserved.
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
135
CSF CHEMOTACTIC ACTIVITY IN MENINGITIS
Fraction
I
Control
24
,
48
I
72
Time ( h o u r s )
number
FIG.2. Sephadex G-75gel filtration of CSF from rabbits with pneumococcal meningitis. Absorbance at 280 nm (-) and chemotactic activity (stippled bars) of each fraction are plotted. Arrows indicate void volume (Vo) and elution position of marker proteins.
FIG. 1. Chemotactic activity in CSF of rabbits with pneumococcal meningitis. Each point represents the mean f SEM of the number of experiments indicated. *Denotes values significantly different from controls (P < .01, t test).
nor C4 were detectable in infected CSF. Absorption of infected CSF or concentrated fractions comprising the 11,000 mol wt peak of chemotactic activity from a Sephadex G-75 column with antihuman C3 and C5 (4) and their CSF contained moderate sera did not reduce chemotactic activity. chemotactic activity after 24 hr but none at Approximately 50 % of the chemotactic activity was removed from an aliquot of 72 hr. infected CSF by absorption with antiserum Chemotactic activity in infected CSF was not dialyzable and was stable at -80" fclr to type I11 pneumococci. Significant chemo2 weeks, 4" for 1 week and 56" for 30 min. tactic activity was detected in filtrates of Activity was abolished by heating at 100" overnight cultures of type 111 pneumococci for 15 min. Supernatants of CSF pre- and in solutions of purified Type 111 pneucipitated with 5 % trichloracetic acid and mococcal capsular polysaccharide. Discussion. This study, showing that granboth supernatants and redissolved precipiulocyte chemotactic activity appears in CSF tates of CSF saturated with ammonium of rabbits with pneumococcal meningitis, sulfate were devoid of chemotactic activity. adds another to the list of chemotacticallyChromatography of infected CSF on active biologic fluids from areas of inflamSephadex G-25 resulted in elution of most of the protein in the exclusion volume (mol mation (8-11). Note that the time course of wt greater than 10,000). Chemotactic ac- appearance of chemotactic activity in CSF tivity was also concentrated in the exclusion during infection closely parallels that of volume, but a small amount of activity was accumulation of the inflammatory mass of present in fractions eluting after the marker granulocytes within the meninges (3). A cause-and-effect relationship between protein insulin, On a Sephadex G-75 column these two observations has not been estab(Fig. 2) most protein again eluted with the lished. Nevertheless, the detection of a exclusion volume (mol wt > 50,000). Howpotential mediator of central nervous system ever, the major peak of chemotactic activity occurred in fractions eluting between ribo- inflammation assumes importance in view nuclease and insulin, corresponding t o a of increasing evidence that meningeal inmol wt of about 11,000; a smaller peak flammation is a determinant of mortality in eluted after insulin, indicating a mol wt of pneumococcal meningitis. If this supposition proves correct in subsequent experiments, approximately 3000. Neither hemolytic complement activity suppression of the inflammatory response
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
136
CSF CHEMOTACTIC ACITVITY IN MENINGITIS
might be beneficial in the treatment of creasing levels for 72 hr after infection. pneumococcal meningitis. Modification of Activity was stable at 56” and was inactivated the chemotactic stimulus is one possible by agents which denature proteins. Gel therapeutic approach t o this problem; char- filtration demonstrated two chemotactically acterization of CSF chemotactic factors active fractions in infected CSF with mol wts and their origin would facilitate this ap- of approximately 3000 and 11,000. Bacterial proac h. products appear to account for a portion of Our preliminary work suggests that CSF the observed CSF chemotactic activity, but chemotactic activity is attributable to rela- the role of host factors remains to be clarified. tively low molecular weight substances which are heat stable and either protein in nature 1. Beaty, H. N., and Oppenheimer, S., N. Eng. J. or dependent upon protein for their activity. Med. 279, 1197 (1968). 2. Sears, M. R., O’Donoghue, J. M., Fisher, H. K., Chemotactic activity of infected CSF was and Beaty, H. N., J. Clin. Invest. 54, 18 (1974). party absorbed by specific pneumococcal 3. McAllister, C. K., O’Donoghue, J. M., and antiserum, suggesting that a bacterial factor Beaty, H. N., J. Infect. Dis. Vol. 132 (1975). accounts for a portion of it. Pneumococci produce low molecular weight ( < 3600) 4. O’Donoghue, J. M., Schweid, A. I., and Beaty, H. N., Proc.Soc.Exp. Biol. Med. 146, 571 (1974). chemotactic activity during growth (12). 5. Gallin, J. I., Clark, R. A., and Kimball, H. R., J. Release of complement derived chemoImmunol. 110, 233 (1973). tactic factors in infected CSF might also be 6. Clark, R. A., and Kimball, H. R., J. Clin. Invest. expected since pneumococci activate com50, 2645 (1971). plement via the alternate pathway (13). 7. Green, A. A. and Hughes, W. L., in “Methods in Although complement activation products Enzymology (S.R. Colowick and N. 0. Kaplan, eds.), Vol. I, p. 67. Academic Press, New York cannot be implicated on the basis of current (1955). evidence, CSF chemotactic activity detected 8. Snyderman, R., Phillips, J. K., and Mergenhagen, in this model has characteristics in common S. E., J. Exp. Med. 134, 1131 (1971). with the potent chemotactic agent C5a. 9. Ward, P. A., and Zwaifler, N. J., J. Clin. Invest. Both are stable at 56” and the mol wt of 50, 606 (1971). rabbit C5a (12,500) (14) is similar to that of 10. Hill, J. H., and Ward, P. A., J. Exp. Med. 133, the major CSF chemotactic peak separated 885 (1971). by gel filtration. Clarification of the roles of 11. Ward, P. A,, and Hill, J. H., J. Immunol. 108, microbial products and host factors in the 1137 (1972). inflammatory response in meningitis is criti- 12. Ward, P. A., Lepow, J. H., and Newman, L. J., Amer. J. Pathol. 52, 725 (1968). cal and deserves further study. Summary. Chemotactic activity was as- 13. Winkelstein, J. A., Shin, H. S., and Wood, W. B., Jr., J. Immunol. 108, 1681 (1972). sayed in CSF of rabbits with pneumococcal 14. Snyderman, R., Phillips, J., and Mergenhagen, meningitis to further characterize the inS. E., Infect. Immunol. 1, 521 (1970). flammatory response in this infection. CSF chemotactic activity was detected in in- Received May 6, 1975. P.S.E.B.M. 1975, Vol. 150.
Downloaded from ebm.sagepub.com at The University of Auckland Library on June 9, 2015
